Development of a Framework for D ata I ntegration, A ssimilation, and L earning for Geological Carbon Sequestration (DIAL-GCS) Project #: DE-FE0026515 Dr. Alex Sun, PhD, PE Bureau of Economic Geology University of Texas at Austin U.S. Department of Energy National Energy Technology Laboratory Mastering the Subsurface Through Technology, Innovation and Collaboration: Carbon Storage and Oil and Natural Gas Technologies Review Meeting August 16-18, 2016
37
Embed
Development of a Framework for Data Integration ... Library/Events/2016/fy16 cs rd/Thur...Development of a Framework for Data Integration, Assimilation, and Learning for ... integrative
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Development of a Framework for Data
Integration, Assimilation, and Learning for
Geological Carbon Sequestration
(DIAL-GCS)Project #: DE-FE0026515
Dr. Alex Sun, PhD, PE
Bureau of Economic Geology
University of Texas at Austin
U.S. Department of Energy
National Energy Technology Laboratory
Mastering the Subsurface Through Technology, Innovation and Collaboration:
Carbon Storage and Oil and Natural Gas Technologies Review Meeting
August 16-18, 2016
Acknowledgements
• DOE/NETL: Brian Dressel
• UT
– BEG: Sue Hovorka (geologist)
• Hoonyoung Jeong (petroleum engineer)
• Katherine Romanak (geochemistry)
• Sergey Fomel, Tieyuan Zhu (geophysics)
• Clay Templeton (web development)
– Information School: David Arctur (data scientist)
Benefit to the Program Carbon storage program goals being addressed
Develop and validate technologies to ensure 99 percent storage permanence
IMS: integrative carbon storage reservoir management technology that combines real-time measurement of reservoir properties with project-specific data management and data processing workflows
Expected benefits of this IMS ProjectTransform scientific knowledge to decision knowledge and public knowledge:
• Promote data sharing and visual analytics– Better collaboration among team members
– Public outreach
• Streamline GCS management and decisionmaking
• Facilitate the optimal allocation of monitoring resources
4
Challenges and Motivations
• GCS is highly interdisciplinary
• Data sharing and discovery is difficult
• Lack of project-specific risk assessment and data
assimilation tools
• Lack of systematic demonstration of machine learning
technologies
5
Project Overview: Goals and Objectives
A. Develop GCS data management module for storing,
querying, exchanging, and visualizing GCS data from
multiple sources and in heterogeneous formats
– Success Criterion: Whether a flexible, user-friendly Web portal
is set up for enabling data exchange and visual analytics
B. Incorporate a complex event processing (CEP) engine
for detecting abnormal situations by seamlessly
combining expert knowledge, rule-based reasoning, and
machine learning
– Success Criterion: Whether a set of decision rules are
developed for identifying abnormal signals in monitoring data
6
Project Overview: Goals and Objectives
C. Enable uncertainty quantification and predictive analytics
using a combination of coupled-process modeling, data
assimilation, and reduced-order modeling
– Success Criterion: Whether a suite of computational tools are
developed for UQ and predictive analytics
D. Integrate and demonstrate the system’s capabilities with
both real and simulated data
– Success Criterion: Whether the IMS tools developed under
Goals A to C are integrated, streamlined, and demonstrated for a
realistic GCS site
7
Technical Status
8
System Architecture of DIAL-GCS
9
Task 2: Sensor Data Schema Development and Serialization (Y1)
Task 3: Development of CEP, machine learning (Y1-2)
Task 4: Coupled Modeling and Data Assimilation (Y1-3)
Task 5: Integration and Demonstration (Y1-3)
Task 2: Sensor Data Management
10
11
DIAL Frontend
12
Web-GIS Admin
A Dashboard for Visual Analytics
13
• Live monitoring datastreams
• Connect to different NoSql servers
• Web-based database query and visualization interface
• Phase 1: Develop “virtual observatories” at for testing monitoring design, risk management, uncertainty quantification
– All models need to make use of high-performance computing
– Main focus: Flow, Seismic, and Geochemistry
• Phase 2: Integrate process-level models with anomaly detection and data assimilation
22
23
• Perform ensemble simulation on TACC
• Send results to server
Modeling
• Compare predictive analytics to observations
• Learning: Detect anomalies
Anomaly Detection • Assimilation: Use
data to improve monitoring design
• Ensemble Kalmanfilter
DA
General Workflow
Flow Modeling
24
Ensemble
Of Realizations
Perform ensemble
simulation
Anomaly detection
UQ, Predictive Analytics, and Machine Learning
By Hoonyoung Jeong
Seismic-Based Anomaly Detection
• Role:
– Virtual observatory for assessing efficacy of
seismic surveys for picking up leakage
– Quantify time-lapse response of seismic
velocity to different leak scenarios or features
• Methodology
– Create 3d velocity and density models using
well logs and reflection seismic data
25
By Tieyuan Zhu & Sergey Fomel
26
Two velocity models (in time)Post-injection after 1 yearPre-injection
27
2D modeling: inline velocity slices
The two slices are windowed from the previous two velocity models.
Pre-injectionPost-injection
after 12 month
28
Simulated seismic data: 2D slices
Difference
Post-injection after 1 year
Pre-injection
Geochemical Data Inversion
• Roles
– Ingesting geochemical data for anomaly
detection
– Leak source attribute recovery
• Methodology
– Adopted a flexible parameterization technique
(level set) for representing source geometry
29
Shallow Aquifer Leak Source Recovery
30
Forward model: TOUGHREACT
Accomplishments to Date
31
Description Task
Developed prototypes for demonstrating datastream management,visualization, and Web-GIS, including capabilities to load data from different sensor sources;
2, 5
Developed multiphysics models, including a multiphase CO2 simulation model, a forward seismic time-lapse response model, and geochemical data inversion capabilities
4
Implemented anomaly detection algorithms and integrated with the Web 3
Summary & Future Work
• Year 1 activities focus on IMS infrastructure building• Developed Web-based protocol for demonstrating all
aspects of IMS
• Our current IMS has user-friendly dashboard and AI capabilities
• Developed modeling capabilities
• In Year 2, we will continue to• Enhance data visualization capability
• Implement more complex event processing algorithms
• Multiphysics model integration
• Develop monitoring network design capabilities
• Build web-based UQ and predictive analysis capabilities
32
Synergy Opportunities– EDX: Big Data computing for data discovery and
anomaly detection
– NRAP: reduced models
– Collaborate with other teams to demonstrate
IMS capabilities
33
Appendix– These slides will not be discussed during the
presentation, but are mandatory
34
Organization Chart
35
Young
BEG
Associate Director
HovorkaSun
(PI)
Romanak
(Co-PI)Yang Xu (TACC) Fomel
Gradudate Students
Arctur (UT Information School)
Gantt Chart
36
Bibliography• Sun, A., Islam, A., Wheeler, M., Identifying Attributes of CO2 Leakage
Zones in Shallow Aquifers Using a Parametric Level Set Method,
submitted to Greenhouse Gases: Science and Technology
• Sun et al., Development of an Intelligent Monitoring System for
Geological Carbon Sequestration (GCS). Abstract submitted to 2016